The preparation of blue light-emitting materials has become a heavily researched field in recent years. This interest is caused by the significant number of potential applications for blue light emitters. The two most commonly used light emitting devices (LED's) are liquid crystal displays (LCDS) and laser diodes. The two most technologically significant applications of blue LED's are electroluminescent displays and read/write heads for optical data storage. Full color electroluminescent red, green and blue (RGB) displays cannot be constructed because they are neither pure green nor pure blue LEDS.
Because the storage density increases inversely to the square of the light source wavelength, a blue LED laser-based optical data storage device, such as a CD-ROM, could store on the order of five times as much data as a standard red LED laser-based CD-ROM. This translates to about 3.25 Gbytes of data for a blue laser based CD, versus 650 Mbytes for the commonly used red laser based CDs. Another less obvious, significant, use of blue (or blue-green) LED's is in trafic signals. Traditional signals utilize a white incandescent light source with an appropriately colored glass filter. The use of an incandescent source, as well as a filter, yields a very inefficient device. The efficiency and longevity of traffic signals can be enhanced through the use of blue, LED's, as well as commonly available red and yellow LEDS. The energy cost savings and the lower maintenance requirements, when multiplied by the number of traffic signals in an average city, represent vast cost reductions. This approach is currently in use in Japan, where blue traffic signals, are used.
GaN and zinc selenide (ZnSe) have emerged as strong candidates for blue light-emitting materials; however, ZnSe suffers from short device lifetimes, relative to GaN. Furthermore, II-VI compounds are rather fragile and are grown at comparatively low temperatures. Thus, work has recently been expanding on GaN materials.
The thermodynamically stable phase of GaN at room temperature and atmospheric pressure is the hexagonal wurzite phase. This material is a direct bandgap semiconductor with a band gap of 3.45 eV. Light emitting diodes require a p-n junction. The fabrication of n-type GaN is not difficult, However, the fabrication of p-type GaN has presented challenges. Zinc has been used as a p-type dopant for GaN, however, its large size and propensity for forming covalent bonds have limited is usefulness. Magnesium has emerged as the dopant of choice for a p-type GaN. The most commonly utilized magnesium sources in the CVD (or MOMBE) growth of GaN:Mg have been organometallic compounds such as bis[cyclopentadienyl]magnesium (Cp.sub.2 Mg), bis[3(dimethyl)propyl]magnesium, and bis[3-(diethylamino)propyl]magnesium. The drawback of these sources is that they create a large carbon impurity (MgC) in the deposited GaN, which destroys the material's electronic usefulness. To compensate for the detrimental effects of the carbon, two approaches have been used. The first is to apply several folds more Mg than needed. This, however, causes lattice deterioration, is expensive, and the devices burn out quickly. The second approach is to attempt to remove the carbon. To counter the formation of MgC, large quantities of H.sub.2 have been injected into the reactor to form CH.sub.4, which vaporizes. However, this in turn causes the formation of MgH.sub.2, which also passivates the material.